Abstract Treatment of EGFR-mutant non-small cell lung cancer (NSCLC) with EGFR-directed tyrosine kinase inhibitors (TKI?s, e.g. gefitinib) has revolutionized thoracic oncology. However, patients develop resistance to gefitinib in two years (on average), and it remains unclear whether gefitinib resistance is pre-existing in a small sub- population of treatment-nave cells, or acquired during therapy. Recently, the Engelman lab published that acquisition of resistance to gefitinib in vitro arises among an epigenetically distinct population of ?persister? cells. This persister state may define the precursors of TKI-refractory cancer, but its molecular features and in vivo significance remain to be determined. The goals of this proposal are to discover the determinants of the persister state, to develop tools to study this state in vivo, and to develop strategies to prevent the accumulation of persister cells. In Aim 1, we will determine whether the persister state occurs in two patient- derived xenograft (PDX) models of EGFR-mutant NSCLC. Both models are TKI-nave, and we estimate that after 3 weeks of gefitinib treatment, the remaining tumor will consist of persister cells. We will perform both RNAseq and ChIP-seq on explanted undrugged or persister tissue, and focus our analysis on whether EMT- related or other epigenetic pathways are upregulated in putative persister cells. Upregulated pathways will be investigated in the drug screen proposed in Aim 3. In Aim 2, we will investigate the role of the Epithelial to Mesenchymal Transition (EMT) in driving the development of a persister state. We have established that EMT occurs in five patient-derived EGFR-mutant cell lines, which express several EMT-related transcription factors. We will determine whether these transcription factors are responsible for gefitinib-induced EMT by engineering knockdown cell lines for each gene. We will then compare both gefitinib-induced EMT and time-to-resistance in the isogenic parental vs. knockdown lines, first in vitro and then in PDXs. In Aim 3, we will screen for compounds that prevent resistance to gefitinib. Our novel platform for 8-week drug screening has identified two candidate compounds (CDK7 and FGFR inhibitors) that lack single-agent efficacy but prevent resistance in combination with gefitinib. We will validate hits in 3D spheroid cultures and in xenograft mouse models, and investigate the mechanism of action of these hits. The work proposed here responds to a paradigm shift in cancer therapy?one in which we are aware that therapy may drive an evolutionary trajectory towards acquired resistance. We will investigate potential mechanisms of this evolution, focusing on EMT. We will explore the chromatin and transcriptional state of persister tissue in PDX models, and will offer a novel drug-screening platform for uniquely long term studies of emerging resistance. This technique may be useful in a variety of cancer types in which acquired resistance limits patient survival.